Seedance 2.5 for your subway singing video.

What can I make with this tool?

A short, single-subject subway singing performance from one photo. The selected version supplies the movement, camera and setting. Your image supplies the adult or pet you want to feature. It is a video-editing workflow with two built-in versions.

Model / supplier

Seedance 2.5 / APIMODELS

Photo

1 adult or pet · JPG, PNG or WebP · up to 10 MiB

Image dimensions

Each edge 300–6000 pixels; width ÷ height 0.4–2.5

Length

Short: about 10 seconds · Extended: about 20 seconds

Quality / file

480p or 720p · vertical MP4, approximately 9:16

Sound

No generated audio or song recording included; add permitted music separately

APIMODELS is the API supplier; Seedance 2.5 is the model. Supplier documentation describes additional capabilities, but this website offers the subway-editing settings above. Read the supplier's model documentation →

Should I choose 10 seconds or 20 seconds?

Choose 10 seconds for a compact reaction or a quick first take; choose 20 seconds when you want the longer performance sequence. Watch both previews before deciding. They are different interpretations of the same evolving trend, so their gestures and timing differ.

Standard 480p uses fewer credits and suits a first attempt. Choose HD 720p when the larger output matters for your edit. A higher output resolution cannot recreate detail missing from the input photo. Exact likeness and clean hand movements can vary at either setting.

How much does Seedance generation cost here?

Free preparation; paid video generation. One-time packs start at $6.99 USD. The price pays for a credit pack, and the chosen version and quality determine how many credits one generation uses.

A shared limit of 3 video requests per rolling 24 hours and one unresolved job per account applies. Buying credits does not reserve a generation slot.

See every pack, credit terms and refund policy →

How do I generate in three steps?

1. Choose your version

   Choose Short for about 10 seconds or Extended for about 20 seconds. The selected preview shows which scene your video will follow.

2. Add your photo and quality

   Use one clear adult or pet photo, JPG, PNG or WebP up to 10 MB. Choose Standard 480p or HD 720p and check the credit cost.

3. Review, then generate

   Confirm the version, approximate length, quality and exact credits together. You can also copy your prompt or save a draft with these settings.

Sign in with Google and have sufficient credits before submitting. A version link selects the requested length in the editor; you still review the quality and cost and confirm the photo-processing consent.

When is each version useful?

• Personal reaction: one adult portrait and the short version keep the performance compact.
• Pet character clip: choose Pet, show both ears and the muzzle, and ask for one gentle paw gesture.
• A longer social edit: use the extended version when you need more of the performance before trimming and adding permitted music.

The tool makes visual performance clips. It does not clone your voice or guarantee word-perfect singing. Read the music guidance.

What should I do when something goes wrong?

The photo is rejected.

Use a real JPG, PNG or WebP within the size, pixel and aspect-ratio limits above. Choose one consenting adult or pet, with a clear face and no extra subjects.

The face changes or the hand looks unusual.

Try a clearer front-facing photo and simplify the gesture. Inspect the whole clip before sharing. A new generation is a separate paid attempt.

The job takes longer than expected.

Check the same job in My drafts. Avoid submitting another while the first outcome is unknown; only one unresolved job per account is admitted.

The request fails.

Confirmed failures return reserved credits. An uncertain provider outcome remains under review with credits reserved until it can be resolved.

I selected 20 seconds but the output is too short.

The service checks output duration against the version saved with your request before marking it completed. A confirmed mismatch fails and releases credits; unverifiable duration remains under review.

The result has no song or the mouth does not match.

This workflow does not include audio. Add permitted music in your publishing app and adjust the timing. Exact lip sync is not guaranteed.

For a stuck job, contact support@aintnomountain.fun and include the job reference from My drafts.

Where do these specifications come from?

The photo limits, version choices and credit costs describe this site's current configuration. Model editing constraints are based on APIMODELS Seedance 2.5 documentation, reviewed 2026-10-23. This guide does not claim a model release date, a benchmark ranking or guaranteed completion time.

Compare this workflow with MakeViral and TrendAI →